[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-948":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":16,"stars30d":16,"stars90d":16,"forks30d":16,"starsTrendScore":16,"compositeScore":17,"rankGlobal":10,"rankLanguage":10,"license":18,"archived":19,"fork":19,"defaultBranch":20,"hasWiki":21,"hasPages":19,"topics":22,"createdAt":10,"pushedAt":10,"updatedAt":23,"readmeContent":24,"aiSummary":25,"trendingCount":16,"starSnapshotCount":16,"syncStatus":26,"lastSyncTime":27,"discoverSource":28},948,"OmegaWiki","skyllwt\u002FOmegaWiki","skyllwt","Karpathy's LLM-Wiki vision, fully realized — wiki-centric full-lifecycle AI research platform powered by Claude Code","",null,"Python",557,83,8,4,0,9.77,"MIT License",false,"main",true,[],"2026-06-12 02:00:21","\u003Cdiv align=\"center\">\n\n\u003Cimg src=\"assets\u002Flogo.png\" width=\"180\" alt=\"ΩmegaWiki Logo\">\n\n# ΩmegaWiki\n\n### Karpathy's LLM-Wiki Vision, Fully Realized\n\n**Your AI Research Platform — From Papers to Publications, Powered by [Claude Code](https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code)**\n\n*From paper ingestion to publication — your research knowledge compounds, never decays.*\n\n[![License: MIT](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-blue.svg)](LICENSE)\n[![Python 3.9+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPython-3.9+-yellow.svg)](https:\u002F\u002Fwww.python.org\u002F)\n[![Skills](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSkills-24-purple.svg)](#skills)\n[![Claude Code](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPowered_by-Claude_Code-d97706.svg)](https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code)\n[![Bilingual](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fi18n-EN_|_中文-orange.svg)](#bilingual-support)\n\n[English](#what-is-ωmegawiki) | [中文](#中文)\n\n\u003C\u002Fdiv>\n\n---\n\n## Team\n\nΩmegaWiki is built by [DAIR Lab](https:\u002F\u002Fcuibinpku.github.io\u002F) at Peking University — a fully agentic platform that automates the complete research pipeline, from knowledge ingestion to paper submission.\n\n\u003Cdiv align=\"center\">\n\u003Ctable>\n  \u003Ctr>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Ca href=\"https:\u002F\u002Fskyllwt.github.io\u002F\">\n        \u003Cimg src=\"assets\u002FWeitongQian_circle.png\" width=\"90\" alt=\"Weitong Qian\"\u002F>\n      \u003C\u002Fa>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Ca href=\"https:\u002F\u002Fskyllwt.github.io\u002F\">\u003Cb>Weitong Qian\u003C\u002Fb>\u003C\u002Fa>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2023\u003C\u002Fsub>\n    \u003C\u002Ftd>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Cimg src=\"assets\u002FBeichengXu_circle.png\" width=\"90\" alt=\"Beicheng Xu\"\u002F>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Cb>Beicheng Xu\u003C\u002Fb>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Ph.D. · 2023\u003C\u002Fsub>\n    \u003C\u002Ftd>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Cimg src=\"assets\u002FZhongaoXie_circle.png\" width=\"90\" alt=\"Zhongao Xie\"\u002F>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Cb>Zhongao Xie\u003C\u002Fb>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2025\u003C\u002Fsub>\n    \u003C\u002Ftd>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Cimg src=\"assets\u002FBowenFan_circle.png\" width=\"90\" alt=\"Bowen Fan\"\u002F>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Cb>Bowen Fan\u003C\u002Fb>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2024\u003C\u002Fsub>\n    \u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Cimg src=\"assets\u002FGuozhengTang_circle.png\" width=\"90\" alt=\"Guozheng Tang\"\u002F>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Cb>Guozheng Tang\u003C\u002Fb>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2024\u003C\u002Fsub>\n    \u003C\u002Ftd>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Ca href=\"https:\u002F\u002Fbrzgw555.github.io\">\n        \u003Cimg src=\"assets\u002FXinzheWu_circle.png\" width=\"90\" alt=\"Xinzhe Wu\"\u002F>\n      \u003C\u002Fa>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Ca href=\"https:\u002F\u002Fbrzgw555.github.io\">\u003Cb>Xinzhe Wu\u003C\u002Fb>\u003C\u002Fa>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2024\u003C\u002Fsub>\n    \u003C\u002Ftd>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Cimg src=\"assets\u002FJialeChen_circle.png\" width=\"90\" alt=\"Jiale Chen\"\u002F>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Cb>Jiale Chen\u003C\u002Fb>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2024\u003C\u002Fsub>\n    \u003C\u002Ftd>\n    \u003Ctd align=\"center\" width=\"165\">\n      \u003Ca href=\"https:\u002F\u002Fmorrowmind.live\">\n        \u003Cimg src=\"assets\u002FMingtianYang_circle.png\" width=\"90\" alt=\"Mingtian Yang\"\u002F>\n      \u003C\u002Fa>\n      \u003Cbr\u002F>\u003Cbr\u002F>\n      \u003Ca href=\"https:\u002F\u002Fmorrowmind.live\">\u003Cb>Mingtian Yang\u003C\u002Fb>\u003C\u002Fa>\n      \u003Cbr\u002F>\n      \u003Csub>PKU\u003C\u002Fsub>\n      \u003Cbr\u002F>\n      \u003Csub>Undergraduate · 2024\u003C\u002Fsub>\n    \u003C\u002Ftd>\n  \u003C\u002Ftr>\n\u003C\u002Ftable>\n\u003C\u002Fdiv>\n\n---\n\n## 🆕 What's New\n\n### 🌐 2026-05-06 · Knowledge Graph Visualization — browser + Obsidian\n\nYour research graph now has two ways to explore:\n\n- **Web UI** — run `python3 tools\u002Fserve.py`, open `http:\u002F\u002Flocalhost:8765\u002F#\u002Fgraph`. Click any node to highlight its neighborhood via BFS, filter by entity type or edge category, double-click to open the full page in the Reader.\n- **Obsidian** — run `\u002Fvisualize --obsidian` to generate a color-coded graph config, or `\u002Fvisualize --canvas` to produce a force-layout Canvas with labeled semantic edges.\n\n### 🔬 2026-05-06 · Methods — Reusable Techniques are Now First-Class\n\nArchitectures, training recipes, evaluation protocols, and other reusable techniques now live in `wiki\u002Fmethods\u002F` as proper wiki entities — with their own pages, source-paper links, and parent\u002Fchild method chains.\n\n---\n\n## What is ΩmegaWiki?\n\nAndrej Karpathy proposed LLM-Wiki: an LLM that **builds and maintains a persistent, structured wiki** from your sources — not a throwaway RAG answer, but compounding knowledge that grows smarter with every paper you feed it.\n\n**ΩmegaWiki takes that idea and runs the full distance.** It's not just a wiki builder — it's a complete research lifecycle platform: from paper ingestion → knowledge graph → gap detection → idea generation → experiment design → paper writing → peer review response. All driven by 24 Claude Code skills, all centered on one wiki as the single source of truth.\n\nDrop your `.tex` \u002F `.pdf` files in a folder. Run one command. Get a fully cross-referenced knowledge base — and then use it to **generate novel research ideas, design experiments, write papers, and respond to reviewers**.\n\n## Why Wiki-Centric, Not RAG?\n\n| | RAG | ΩmegaWiki |\n|---|---|---|\n| **Knowledge persistence** | Rediscovered on every query | Compiled once, maintained forever |\n| **Structure** | Flat chunk store | 9 typed entities with relationships |\n| **Cross-references** | None — chunks are isolated | Bidirectional wikilinks + typed graph |\n| **Knowledge gaps** | Invisible | Explicitly tracked, drive research |\n| **Failed experiments** | Lost | First-class anti-repetition memory |\n| **Output** | Chat answers | Papers, surveys, experiment plans, rebuttals |\n| **Compounding** | No — same cost every query | Yes — each paper enriches the whole graph |\n\n## Architecture\n\n\u003Cdiv align=\"center\">\n\u003Cimg src=\"assets\u002Farchitecture.png\" width=\"700\" alt=\"ΩmegaWiki Architecture\">\n\u003C\u002Fdiv>\n\nEvery skill reads from and writes back to the wiki. Knowledge compounds — each new paper enriches the whole graph. Failed experiments aren't discarded; they become anti-repetition memory that prevents re-exploring dead ends.\n\n## Quick Start\n\n**Prerequisites:** Python 3.9+, Node.js 18+\n\n```bash\n# 1. Clone\ngit clone https:\u002F\u002Fgithub.com\u002Fskyllwt\u002FOmegaWiki.git\ncd OmegaWiki\n\n# 2. Install Claude Code\nnpm install -g @anthropic-ai\u002Fclaude-code\nclaude login\n\n# 3. One-click setup\nchmod +x setup.sh && .\u002Fsetup.sh        # Linux \u002F macOS\n# Windows (PowerShell):\n#   powershell -ExecutionPolicy Bypass -File .\\setup.ps1\n# setup creates .venv for OmegaWiki\n# the script does not keep your shell activated, but \u002Finit will use .venv automatically\n\n# 4. Put your own papers in raw\u002Fpapers\u002F (.tex or .pdf)\n#    Optional: add intent notes to raw\u002Fnotes\u002F and saved pages to raw\u002Fweb\u002F\n#    \u002Finit and direct local \u002Fingest will manage generated inputs under raw\u002Fdiscovered\u002F and raw\u002Ftmp\u002F\n\n# 5. Build your wiki\nclaude\n# Then type: \u002Finit [your-research-topic]\n```\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>Manual setup (Linux \u002F macOS)\u003C\u002Fb>\u003C\u002Fsummary>\n\n```bash\npython3 -m venv .venv && source .venv\u002Fbin\u002Factivate\npip install -r requirements.txt\ncp .env.example .env                 # Edit to add API keys\ncp config\u002Fsettings.local.json.example .claude\u002Fsettings.local.json\n```\n\n\u003C\u002Fdetails>\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>Manual setup (Windows \u002F PowerShell)\u003C\u002Fb>\u003C\u002Fsummary>\n\n```powershell\npython -m venv .venv\n.\\.venv\\Scripts\\Activate.ps1\npip install -r requirements.txt\nCopy-Item .env.example .env          # Edit to add API keys\nCopy-Item config\\settings.local.json.example .claude\\settings.local.json\n```\n\nNote: native Windows is supported for the local pipeline. Remote-GPU\nexperiments via `\u002Fexp-run --env remote` rely on `ssh`\u002F`rsync`\u002F`screen`\nand are best run from WSL2 or Linux\u002FmacOS.\n\n\u003C\u002Fdetails>\n\n### API Keys\n\n| Key | Required? | How to get | What it enables |\n|-----|-----------|-----------|-----------------|\n| `ANTHROPIC_API_KEY` | **Yes** | `claude login` (automatic) | Powers all Claude Code skills |\n| `SEMANTIC_SCHOLAR_API_KEY` | Optional | [semanticscholar.org\u002Fproduct\u002Fapi](https:\u002F\u002Fwww.semanticscholar.org\u002Fproduct\u002Fapi) (free) | Citation graph, paper search |\n| `DEEPXIV_TOKEN` | Optional | `setup.sh` auto-registers | Semantic search, TLDR, trending |\n| `LLM_API_KEY` + `LLM_BASE_URL` + `LLM_MODEL` | Optional | Any OpenAI-compatible API | Cross-model review |\n\n> **Cross-model review**: ΩmegaWiki uses a second LLM as an independent reviewer for ideas, experiments, and paper drafts. Works with **any OpenAI-compatible API** — DeepSeek, OpenAI, Qwen, OpenRouter, SiliconFlow, etc. If not configured, skills still work in Claude-only mode.\n\n## Skills\n\n24 slash commands spanning the full research lifecycle:\n\n### Phase 0: Setup\n\n| Command | What it does |\n|---------|-------------|\n| `\u002Fsetup` | First-time configuration (API keys, language, dependencies) |\n| `\u002Freset \u003Cscope>` | Destructive cleanup: `wiki \\| raw \\| log \\| checkpoints \\| all` |\n\n### Phase 1: Knowledge Foundation\n\n| Command | What it does |\n|---------|-------------|\n| `\u002Fprefill \u003Cdomain>` | Optionally seed `foundations\u002F` with background knowledge |\n| `\u002Finit [topic]` | Bootstrap a full wiki from user raw sources plus optional discovery |\n| `\u002Fingest \u003Csource>` | Parse a paper → wiki pages + cross-references |\n| `\u002Fdiscover` | Recommend ranked next-read papers from anchors, a topic, or the current wiki |\n| `\u002Fedit \u003Crequest>` | Add\u002Fremove sources or update wiki content |\n| `\u002Fask \u003Cquestion>` | Query the wiki, crystallize answers back |\n| `\u002Fcheck` | Health scan: broken links, missing cross-refs, consistency |\n\n### Phase 2: Research Pipeline\n\n| Command | What it does |\n|---------|-------------|\n| `\u002Fdaily-arxiv` | Auto-fetch & filter new arXiv papers (+ GitHub Actions cron) |\n| `\u002Fideate` | Multi-phase idea generation from cross-topic connections |\n| `\u002Fnovelty \u003Cidea>` | Multi-source novelty verification (web + S2 + wiki + review LLM) |\n| `\u002Freview \u003Cartifact>` | Cross-model adversarial review for any research artifact |\n| `\u002Fexp-design \u003Cidea>` | Idea-driven experiment + ablation design |\n| `\u002Fexp-run \u003Cexperiment>` | Implement + deploy + monitor (local or remote GPU) |\n| `\u002Fexp-status` | Dashboard for running experiments; auto-collect results |\n| `\u002Fexp-eval \u003Cexperiment>` | Verdict gate → auto-update the linked idea + graph |\n| `\u002Frefine \u003Cartifact>` | Multi-round: produce → review → fix → re-review |\n\n### Phase 3: Writing & Submission\n\n| Command | What it does |\n|---------|-------------|\n| `\u002Fsurvey` | Generate Related Work from wiki knowledge |\n| `\u002Fpaper-plan \u003Cideas>` | Outline from validated-idea graph + evidence matrix |\n| `\u002Fpaper-draft \u003Cplan>` | Draft LaTeX + figures, section by section |\n| `\u002Fpaper-compile \u003Cdir>` | Compile → PDF, auto-fix, verify page\u002Fanonymity |\n| `\u002Fresearch \u003Cdirection>` | End-to-end orchestrator with human gates |\n| `\u002Frebuttal \u003Creviews>` | Parse reviewer comments → draft point-by-point responses |\n\n## Wiki Structure\n\n### 9 Entity Types\n\n| Type | Directory | Purpose |\n|------|-----------|---------|\n| **Paper** | `papers\u002F` | Structured summary: problem\u002Fkey idea\u002Fmethod\u002Fexperiment+results\u002Flimitations + tldr\u002Fcontribution_type\u002Fdatasets |\n| **Concept** | `concepts\u002F` | Cross-paper technical concept with variants, comparisons, definition, linked ideas |\n| **Topic** | `topics\u002F` | Research direction map with SOTA tracker, key benchmarks, and open problems (split into known + methodological gaps) |\n| **Person** | `people\u002F` | Researcher profile with research areas, recent work, and a researcher\u002Fteam\u002Forganization type |\n| **Idea** | `ideas\u002F` | Research idea with lifecycle, novelty argument & score, target venue |\n| **Experiment** | `experiments\u002F` | Full record: hypothesis → setup → results → updates to the linked idea |\n| **Method** | `methods\u002F` | Reusable, citable technique entity (cross-paper); links to source papers and parent\u002Fchild methods |\n| **Summary** | `Summary\u002F` | Domain-wide survey across topics |\n| **Foundation** | `foundations\u002F` | Background knowledge (terminal: receives inward links, writes none) |\n\n### Knowledge Graph\n\nSemantic relationships are stored in `graph\u002Fedges.jsonl`; bibliographic paper citations are stored separately in `graph\u002Fcitations.jsonl`.\n\nPaper-paper semantic edges include `same_problem_as`, `similar_method_to`, `complementary_to`, `builds_on`, `compares_against`, `improves_on`, `challenges`, and `surveys`. Paper-concept edges use `introduces_concept`, `uses_concept`, `extends_concept`, and `critiques_concept`. Workflow edges (`supports`, `contradicts`, `tested_by`, `invalidates`, `addresses_gap`, `inspired_by`, `derived_from`) span experiments, ideas, methods, and concepts.\n\nAll pages use **Obsidian `[[wikilink]]` format** — open `wiki\u002F` in Obsidian for visual graph exploration.\n\n## Automation\n\n**GitHub Actions** runs `\u002Fdaily-arxiv` at UTC 00:00 daily:\n\n1. Add `ANTHROPIC_API_KEY` to repo **Settings → Secrets**\n2. `.github\u002Fworkflows\u002Fdaily-arxiv.yml` fetches arXiv, runs ingestion, auto-commits\n\n## Project Structure\n\n```\nOmegaWiki\u002F\n├── CLAUDE.md                    # Runtime schema & rules\n├── wiki\u002F                        # Knowledge base (LLM-maintained)\n│   ├── papers\u002F                  #   Structured paper summaries\n│   ├── concepts\u002F                #   Cross-paper technical concepts\n│   ├── topics\u002F                  #   Research direction maps\n│   ├── people\u002F                  #   Researcher profiles\n│   ├── ideas\u002F                   #   Research ideas (with lifecycle)\n│   ├── experiments\u002F             #   Experiment records\n│   ├── methods\u002F                 #   Reusable cross-paper method entities\n│   ├── Summary\u002F                 #   Domain-wide surveys\n│   ├── foundations\u002F             #   Background knowledge (terminal pages)\n│   ├── outputs\u002F                 #   Generated artifacts\n│   ├── graph\u002F                   #   Auto-generated: edges, context, gaps\n│   ├── index.md                 #   Content catalog\n│   └── log.md                   #   Chronological log\n├── raw\u002F                         # Source materials\n│   ├── papers\u002F                  #   User-owned .tex \u002F .pdf files\n│   ├── discovered\u002F              #   \u002Finit and \u002Fdaily-arxiv-downloaded external papers\n│   ├── tmp\u002F                     #   generated prepared local sidecars for \u002Finit and direct local \u002Fingest\n│   ├── notes\u002F                   #   User-owned .md notes\n│   └── web\u002F                     #   User-owned HTML \u002F Markdown\n├── tools\u002F                       # Deterministic Python helpers\n│   ├── research_wiki.py         #   Wiki engine (20 CLI commands)\n│   ├── init_discovery.py        #   \u002Finit prepare + plan + fetch helper\n│   ├── discover.py              #   \u002Fdiscover candidate gathering, dedup, ranking\n│   ├── lint.py                  #   Structural validation (10 checks)\n│   ├── reset_wiki.py            #   Scoped destructive cleanup helper\n│   ├── fetch_arxiv.py           #   arXiv RSS fetcher\n│   ├── fetch_s2.py              #   Semantic Scholar API\n│   ├── fetch_deepxiv.py         #   DeepXiv semantic search\n│   ├── fetch_wikipedia.py       #   Wikipedia fetcher (used by \u002Fprefill)\n│   └── remote.py                #   SSH ops for remote experiments\n├── .claude\u002Fskills\u002F              # 24 Claude Code skill definitions\n├── i18n\u002F                        # Bilingual: en\u002F (canonical) + zh\u002F\n├── config\u002F                      # Configuration templates\n├── mcp-servers\u002F                 # Cross-model review server\n└── .github\u002Fworkflows\u002F           # Daily arXiv cron\n```\n\n\n## Bilingual Support\n\nΩmegaWiki ships in English and Chinese:\n\n```bash\n.\u002Fsetup.sh --lang en   # English (default)\n.\u002Fsetup.sh --lang zh   # 中文\n```\n\n---\n\n## Contributing\n\nWe welcome contributions! See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n## LLM API Configuration \u002F 大模型 API 配置\n\nΩmegaWiki runs on **Claude Code**, which speaks the **Anthropic API** protocol. You can use Claude directly, or route Claude Code to any third-party provider that exposes an Anthropic-compatible endpoint by overriding a few environment variables.\n\nΩmegaWiki 基于 **Claude Code**,Claude Code 使用 **Anthropic API** 协议通信。你既可以直接使用 Claude,也可以通过覆盖几个环境变量,把 Claude Code 指向任意支持 Anthropic 协议的第三方供应商。\n\n### Option A — Native Claude \u002F 原生 Claude\n\n```bash\nclaude login   # OAuth, no manual config \u002F OAuth 登录,无需手动配置\n```\n\n### Option B — Third-party Anthropic-compatible API \u002F 第三方 Anthropic 兼容 API\n\nPick a provider below, paste the snippet into `~\u002F.claude\u002Fsettings.json` (or the project's `.claude\u002Fsettings.json`), and replace the `\u003C...>` placeholder with your own API key. Model names and extra options are taken from each provider's official Claude Code docs — if anything stops working (e.g. a model is renamed), check the provider's website.\n\n从下方任选一个供应商,把对应配置粘贴到 `~\u002F.claude\u002Fsettings.json`(或项目的 `.claude\u002Fsettings.json`),并把 `\u003C...>` 占位符替换为你自己的 API key。模型名与额外选项均来自各供应商官方 Claude Code 文档;若出现问题(例如模型改名),请查询对应官网。\n\n#### MiMo (小米)\n\n```json\n{\n  \"env\": {\n    \"ANTHROPIC_BASE_URL\": \"https:\u002F\u002Fapi.xiaomimimo.com\u002Fanthropic\",\n    \"ANTHROPIC_AUTH_TOKEN\": \"\u003Cyour-mimo-key>\",\n    \"ANTHROPIC_MODEL\": \"mimo-v2.5\",\n    \"ANTHROPIC_DEFAULT_SONNET_MODEL\": \"mimo-v2.5\",\n    \"ANTHROPIC_DEFAULT_OPUS_MODEL\": \"mimo-v2.5-pro\",\n    \"ANTHROPIC_DEFAULT_HAIKU_MODEL\": \"mimo-v2.5\"\n  }\n}\n```\n\n#### DeepSeek\n\n```json\n{\n  \"env\": {\n    \"ANTHROPIC_BASE_URL\": \"https:\u002F\u002Fapi.deepseek.com\u002Fanthropic\",\n    \"ANTHROPIC_AUTH_TOKEN\": \"\u003Cyour-deepseek-key>\",\n    \"ANTHROPIC_MODEL\": \"deepseek-v4-pro[1m]\",\n    \"ANTHROPIC_DEFAULT_OPUS_MODEL\": \"deepseek-v4-pro[1m]\",\n    \"ANTHROPIC_DEFAULT_SONNET_MODEL\": \"deepseek-v4-pro[1m]\",\n    \"ANTHROPIC_DEFAULT_HAIKU_MODEL\": \"deepseek-v4-flash\",\n    \"CLAUDE_CODE_SUBAGENT_MODEL\": \"deepseek-v4-flash\",\n    \"CLAUDE_CODE_EFFORT_LEVEL\": \"max\"\n  }\n}\n```\n\n#### Kimi (Moonshot)\n\n```json\n{\n  \"env\": {\n    \"ANTHROPIC_BASE_URL\": \"https:\u002F\u002Fapi.moonshot.ai\u002Fanthropic\",\n    \"ANTHROPIC_AUTH_TOKEN\": \"\u003Cyour-moonshot-key>\",\n    \"ANTHROPIC_MODEL\": \"kimi-k2.5\",\n    \"ANTHROPIC_DEFAULT_OPUS_MODEL\": \"kimi-k2.5\",\n    \"ANTHROPIC_DEFAULT_SONNET_MODEL\": \"kimi-k2.5\",\n    \"ANTHROPIC_DEFAULT_HAIKU_MODEL\": \"kimi-k2.5\",\n    \"CLAUDE_CODE_SUBAGENT_MODEL\": \"kimi-k2.5\",\n    \"ENABLE_TOOL_SEARCH\": \"false\"\n  }\n}\n```\n\n#### GLM (Z.AI)\n\n```json\n{\n  \"env\": {\n    \"ANTHROPIC_BASE_URL\": \"https:\u002F\u002Fapi.z.ai\u002Fapi\u002Fanthropic\",\n    \"ANTHROPIC_AUTH_TOKEN\": \"\u003Cyour-zai-key>\",\n    \"API_TIMEOUT_MS\": \"3000000\"\n  }\n}\n```\n\n> Z.AI applies a default server-side model mapping, so no explicit `ANTHROPIC_MODEL` is needed.\n> Z.AI 默认在服务端做模型映射,无需显式设置 `ANTHROPIC_MODEL`。\n\n**Skip the Claude Code onboarding** \u002F **跳过 Claude Code 初始引导**\n\nWhen using a third-party key (instead of `claude login`), Claude Code's first-run onboarding won't complete automatically. Create or edit `.claude.json` and mark it done:\n\n使用第三方 key 时不会走 `claude login`,Claude Code 首次启动的引导不会自动完成。创建或编辑 `.claude.json`,手动标记引导已完成:\n\n- macOS \u002F Linux: `~\u002F.claude.json`\n- Windows: `\u003Cuser-home>\\.claude.json`\n\n```json\n{\n  \"hasCompletedOnboarding\": true\n}\n```\n\nThen run `claude` as usual. \u002F 保存后正常运行 `claude` 即可。\n\n---\n\n## Community \u002F 交流群\n\n\u003Cimg src=\"assets\u002Fwechat_group.png\" width=\"240\" alt=\"WeChat Group QR Code\">\n\nScan to join the ΩmegaWiki WeChat group \u002F 扫码加入微信交流群\n\n## Acknowledgments\n\n- **Andrej Karpathy** — for the LLM-Wiki concept that inspired this project\n- **[Claude Code](https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code)** — the AI agent runtime that powers ΩmegaWiki\n\n## Star History\n\n[![Star History Chart](https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=skyllwt\u002FOmegaWiki&type=Date)](https:\u002F\u002Fstar-history.com\u002F#skyllwt\u002FOmegaWiki&Date)\n\n## License\n\n[MIT](LICENSE) — use it, fork it, build on it.\n\n---\n\n## 中文\n\n### ΩmegaWiki 是什么？\n\nAndrej Karpathy 提出了 LLM-Wiki 概念：让 LLM **构建并维护一个持久的、结构化的 wiki**，而不是一次性的 RAG 回答。知识持续积累，每一篇新论文都让整个知识图谱更强。\n\n**ΩmegaWiki 将这个理念完整实现。** 它不仅是 wiki 构建器，更是完整的研究全流程平台：从论文摄入 → 知识图谱 → 缺口检测 → 想法生成 → 实验设计 → 论文写作 → 同行评审回复。24 个 Claude Code Skills 驱动，一个 wiki 作为唯一的知识中枢。\n\n### 为什么选择 Wiki 而不是 RAG？\n\n| | RAG | ΩmegaWiki |\n|---|---|---|\n| **知识持久性** | 每次查询都重新发现 | 编译一次，持续维护 |\n| **结构** | 扁平的 chunk 存储 | 9 种实体类型 + 关系图 |\n| **交叉引用** | 无 — chunk 彼此孤立 | 双向 wikilink + 类型化边 |\n| **知识缺口** | 不可见 | 显式追踪，驱动研究方向 |\n| **失败实验** | 丢失 | 一等公民，防止重复探索 |\n| **输出** | 聊天回答 | 论文、综述、实验方案、审稿回复 |\n| **复利效应** | 无 — 每次查询成本相同 | 有 — 每篇论文丰富整个图谱 |\n\n### 快速开始\n\n**前置条件：** Python 3.9+, Node.js 18+\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fskyllwt\u002FOmegaWiki.git && cd OmegaWiki\n\n# 安装 Claude Code\nnpm install -g @anthropic-ai\u002Fclaude-code\nclaude login\n\n# 一键配置\nchmod +x setup.sh && .\u002Fsetup.sh --lang zh        # Linux \u002F macOS\n# Windows (PowerShell):\n#   powershell -ExecutionPolicy Bypass -File .\\setup.ps1 -Lang zh\n# setup 会为 OmegaWiki 创建 .venv\n# 脚本不会把你当前 shell 永久激活，但 \u002Finit 会自动使用 .venv\n\n# 把你自己的论文放入 raw\u002Fpapers\u002F（.tex 或 .pdf）\n# 可选：把意图笔记放入 raw\u002Fnotes\u002F，网页存档放入 raw\u002Fweb\u002F\n# \u002Finit 与直接本地 \u002Fingest 会自动管理 raw\u002Fdiscovered\u002F 与 raw\u002Ftmp\u002F 下的生成内容\n# 启动 Claude Code\nclaude\n# 输入：\u002Finit [你的研究方向]\n```\n\n> **Windows 用户**：本地 pipeline 已原生支持。`\u002Fexp-run --env remote` 远程 GPU 实验依赖 `ssh`\u002F`rsync`\u002F`screen`，建议在 WSL2 或 Linux\u002FmacOS 下运行。\n\n### API Key 说明\n\n| Key | 必须？ | 获取方式 | 用途 |\n|-----|--------|---------|------|\n| `ANTHROPIC_API_KEY` | **是** | `claude login` | 驱动所有 Skill |\n| `SEMANTIC_SCHOLAR_API_KEY` | 可选 | [semanticscholar.org](https:\u002F\u002Fwww.semanticscholar.org\u002Fproduct\u002Fapi)（免费） | 引用图谱、论文搜索 |\n| `DEEPXIV_TOKEN` | 可选 | `setup.sh` 自动注册 | 语义搜索、热门趋势 |\n| `LLM_API_KEY` + `LLM_BASE_URL` + `LLM_MODEL` | 可选 | 任意 OpenAI 兼容 API | 跨模型评审 |\n\n### 24 个 Skill 命令\n\n| 命令 | 功能 |\n|------|------|\n| `\u002Fsetup` | 首次配置（API key、语言、依赖） |\n| `\u002Freset` | 按范围销毁性清理：`wiki \\| raw \\| log \\| checkpoints \\| all` |\n| `\u002Fprefill` | 可选地预填 `foundations\u002F` 背景知识 |\n| `\u002Finit` | 基于用户 raw 素材并按需做外部发现来搭建 wiki |\n| `\u002Fingest` | 消化论文，创建页面 + 交叉引用 |\n| `\u002Fdiscover` | 从 anchor、topic 或当前 wiki 推荐排序后的下一批待读论文 |\n| `\u002Fedit` | 增删 raw 或更新 wiki |\n| `\u002Fask` | 对 wiki 提问 |\n| `\u002Fcheck` | wiki 健康检查 |\n| `\u002Fdaily-arxiv` | 每日 arXiv 新论文（CI 自动） |\n| `\u002Fideate` | 跨方向构思研究 idea |\n| `\u002Fnovelty` | 多源新颖性验证 |\n| `\u002Freview` | 跨模型评审 |\n| `\u002Fexp-design` | idea 驱动的实验设计 |\n| `\u002Fexp-run` | 部署 + 监控实验 |\n| `\u002Fexp-status` | 实验状态看板 |\n| `\u002Fexp-eval` | 裁决 → 自动更新关联 idea |\n| `\u002Frefine` | 多轮迭代改进 |\n| `\u002Fsurvey` | 生成 Related Work |\n| `\u002Fpaper-plan` | idea 图谱 + 实验证据 → 论文提纲 |\n| `\u002Fpaper-draft` | 提纲 + wiki → LaTeX 草稿 |\n| `\u002Fpaper-compile` | 编译 → PDF，自动修复 |\n| `\u002Fresearch` | 端到端研究编排器 |\n| `\u002Frebuttal` | 解析评审意见 → 逐条回复 |\n\n---\n\n\u003Cdiv align=\"center\">\n\n**Built with [Claude Code](https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code)**\n\nIf this project helps your research, give it a ⭐\n\n\u003C\u002Fdiv>\n","ΩmegaWiki 是一个基于 Claude Code 的全生命周期 AI 研究平台，旨在实现 Karpathy 的 LLM-Wiki 愿景。它通过自动化从论文阅读到发表的整个研究流程，帮助研究人员高效管理知识。项目使用 Python 开发，具备强大的文档处理和自然语言处理能力，支持多语言界面。适合需要系统化整理文献资料、追踪研究进展及撰写学术文章的研究人员或团队使用。",2,"2026-05-06 17:29:10","CREATED_QUERY"]